摘 要:为了提升产品包装缺陷识别的速度,文章结合机器视觉技术,设计一种信息型智能包装缺陷识别方法。通过提取包装缺陷特征,设计多网格缺陷识别结构,以此构建机器视觉双向缺陷识别模型,采用多帧处理法获取缺陷识别结果。测试结果显示,该文所设计的机器视觉包装缺陷测试组最终得出的定向识别时间均控制在 1.1 s 以内,表明在机器视觉技术的辅助之下,对产品包装缺陷的识别效果更具有实际的应用价值。
关键词:机器视觉;信息型;智能包装;缺陷识别;识别方法;包装检测
DOI:10.19850/j.cnki.2096-4706.2022.20.038
中图分类号:TP391.4 文献标识码:A 文章编号:2096-4706(2022)20-0163-03
An Information-based Intelligent Packaging Defect Recognition Method Based on Machine Vision
LYU Ruijie, SUN Xiangming
(Luanzhou Jihong Packaging Co., Ltd., Tangshan 063700, China)
Abstract: In order to improve the speed of product packaging defect recognition, this paper combines machine vision technology to design an information-based intelligent packaging defect recognition method. By extracting packaging defect features, a multi grid defect recognition structure is designed to build a machine vision bidirectional defect recognition model, and the defect recognition results are obtained by using multi frame processing method. The test results show that the orientation recognition time finally obtained by the machine vision packaging defect test group designed in this paper is controlled within 1.1 s. It shows that with the help of machine vision technology, the recognition effect of product packaging defects has more practical application value.
Keywords: machine vision; information-based; intelligent packaging; defect recognition; recognition method; packaging testing
参考文献:
[1] 陈宗阳,赵辉,吕永胜,等 . 基于改进 MobileNetV2 网络的涂层表面缺陷识别方法 [J]. 哈尔滨工程大学学报,2022,43(4):572-579.
[2] 张赛,王应彪,杨谭,等 . 基于改进 LeNet-5 模型的木材表面典型缺陷识别方法研究 [J]. 木材科学与技术,2021,35(6):31-37.
[3] 潘铭志,潘宏侠,许昕,等 . 复杂结构装配质量检测与缺陷识别的频响特性方法研究 [J].火炮发射与控制学报,2021,42(3):87-92+99.
[4] 刘雷霆,袁芬.机器视觉技术的包装缺陷在线识别系统 [J].国外电子测量技术,2021,40(5):109-113.
[5] 陈轶楠,葛斌,王俊,等 . 基于集成分类器的泡罩包装药品缺陷识别 [J]. 包装工程,2021,42(1):250-259.
[6] 李建明,杨挺,王惠栋 . 基于深度学习的工业自动化包装缺陷检测方法 [J]. 包装工程,2020,41(7):175-184.
[7] 苏东敏,陈友兴,王召巴,等 . 基于时频分析的 CFRP 分层缺陷识别研究 [J]. 测试技术学报,2022,36(3):185-189.
作者简介:吕瑞杰(1980.08—),男,汉族,河北邯郸人,总经理,研究方向:智能包装设计及工艺研究;孙相明(1997.08—),男,汉族,河北邢台人,工程师,本科,研究方向:智能包装设计及工艺研究。